Ten Lines of Code and a Reddit Bot

On September 29, 2025, a Tulsa company, Greystone Logistics, put out a press release with an unusual headline. The company said, per its release, that it was sharing an “independent analysis” by something called BeyondSPX Research, published on September 4. Its chief executive, Warren Kruger, was quoted saying the company’s “historical efforts and current direction are well described” by the report. The release did not say who ran BeyondSPX, or how old he was.

According to coverage of a later Wall Street Journal report, he was 15.

The arc

The WSJ piece itself sits behind a paywall that this publication did not get through, so what follows rests on three outlets that summarized it, plus the company’s own site. VnExpress International reports that Nick Dobroshinsky is a high school student and one of three teen founders in the Journal’s report. The idea, VnExpress says, took shape toward the end of eighth grade; it quotes him telling the WSJ, “Around the end of eighth grade, I wanted to make some sort of business.”

The product is a research site that generates reports on small- and mid-cap public companies, the kind with little or no analyst coverage. The company’s own About page says he began value investing in seventh grade, built a research platform he used himself, and found that other investors valued it too. 36Kr, in a German-language version, says he founded it in July 2024, at 14, and that it covers more than 5,000 US stocks including OTC names. VnExpress and Caliber.az both give the figure of more than 50,000 monthly users and say it was free, with monetization planned. Neither says how that was measured. 36Kr adds a 250,000 total-investor figure that the other two do not repeat, and this piece does not rely on it. No source gives a launch date beyond 36Kr’s, so treat July 2024 as one outlet’s account.

In 2026, per the site, BeyondSPX was renamed EveryTicker, and the company now lists a San Francisco headquarters. Full reports now require a sign-in.

How it was built, and sold

The detail that traveled was the code. VnExpress says he told the Journal he had written only about ten lines himself, relying on Anthropic’s Claude, with ChatGPT and Gemini alongside, to design the architecture. 36Kr describes the same split: Claude to generate the software, other models for content. It is a self-description relayed through translation, and “ten lines” is best read as a boast about the proportion, not an audit of the repository.

Distribution was a bot. VnExpress says that rather than hire a team, he used Reddit bots to promote the site in investing threads, and renders the bot’s pitch as: “There’s a bunch of investing tools and BeyondSPX is one of them.” Caliber.az and 36Kr describe the same bot, triggered when someone asks for good investing tools. None of the coverage says which subreddits were involved or how the moderators of those communities felt about it. That omission is not small. Automated comments written to look like organic advice are the thing many forums exist to keep out.

The Greystone problem

The most interesting fact is the press release. According to 36Kr, an investor-relations consultant, Brandon Hopkins, did not initially know the founder was 15; the German text paraphrases him as suggesting the author might have some form of mass AI analysis. The release itself contains no disclosure of whether BeyondSPX was paid or had any arrangement with Greystone, and it was issued by Greystone, the company being analyzed. The distribution page for it carries only a third-party-content notice. That is not evidence of anything improper. It is the absence of information a reader would want.

What this is evidence of

Three things, none of them settled.

First, vibe coding has moved from demo to product. A founder who steers models and does not write code can ship something that real users return to, and the age of the person doing the steering matters less than it used to. Second, the one-person company now has a credible distribution tool in the same technology. A bot costs nothing per comment, which is exactly why platforms treat it as a threat. Third, and the real risk: machine-written analysis of thinly covered stocks is cheap, fluent and hard to check. Micro-cap and OTC companies often have little coverage because there is little to cover. A report that reads like research can be quoted in a press release by the company it describes, and an investor skimming that release may not know the analysis was generated, who prompted it, or whether anyone fact-checked the numbers.

The caveats cut both ways. Luck is visible here: Claude and its peers became good enough at long-form summaries at roughly the time a teenager with an investing habit wanted one. The user count is self-reported and unaudited. Whether free traffic converts to paying customers is, by the founder’s own plan, still ahead of him. And the company has since changed its name, which the sources do not explain.

What the story shows is not that a teenager can beat analysts. It is that the cost of producing something that looks like analysis has collapsed, and the cost of checking it has not.

Not financial advice.